Researchers at Helmholtz Munich, the Ludwig Maximilians University Munich (LMU) and collaborating institutions have developed an artificial intelligence (AI) framework that maps disease-related changes throughout the entire mouse body at cellular resolution. Using the new platform, called MouseMapper, the researchers uncovered widespread inflammation and previously unrecognized damage to facial sensory nerves caused by obesity. Importantly, they also identified corresponding molecular signatures in human tissue, suggesting that key features of obesity-associated nerve damage are conserved across species. The findings are published today in the journal Nature.

The University of North Florida has received a National Science Foundation award that will allow researchers to make 3D‑printed metal parts more reliable and less wasteful.

Once only achievable in the far-fetched imaginations of science fiction writers, 3D printing has gone mainstream. Relatively inexpensive machines allow individuals to design and print everything from board games and desk accessories to replacement parts for household appliances and more. One of the biggest selling points is the ability to recycle a printed piece into something new, offering a potential pathway to more sustainable living. However, the highest quality 3D printing used to create extremely precise structures at the submicron scale — a method called stereolithography — uses process that involves photocuring resin by exposing the material to ultraviolet (UV) light. This irreversibly chemically changes the resin, making it impossible to recycle.

A new study published in Big Earth Data systematically evaluates popular Uncertainty Quantification (UQ) methods and metrics for AI/ML-based geospatial applications, addressing challenges of predictive uncertainty, trustworthiness, and real-time integration in complex Earth system modelling. Using a PM2.5 calibration case study, it demonstrates that Deep Ensembles and Bayesian Neural Networks implemented in TensorFlow achieved the best reliability and calibration performance, while also highlighting framework-specific differences between TensorFlow and PyTorch in geospatial UQ applications.

A University of Florida engineering professor says commercially available tools used to detect AI-generated text in scientific literature are ineffective.

Researchers have developed a new method to overcome long-standing challenges in manipulating nanowires—tiny, rod-shaped structures essential for next-generation electronics, quantum devices, and biological sensors.

Three faculty researchers from The University of Texas at San Antonio were celebrated with the inaugural Texas Innovation Award, a new statewide award recognizing leadership in translating academic research into real-world impact.